Portable equipment for retina diagnosis controlled by smartphone

Abstract

Millions of people lose their sight every year on our planet. According to the Brazilian Ophthalmological Census (CBO 2014), four million people have serious visual impairments, a hundred thousand being children. In addition, 23% of the population has no access to ophthalmologists and about a half of the cases of blindness are associated with retinal diseases. With the prospect of offering a solution that contributes to this strong social demand, we present this project of a hand-held portable fundus camera, which can be coupled and controlled by smartphones, being compact, light and more affordable compared to the current table-top fundus cameras. We are proposing the creation of a new optical architecture for retina illumination and imaging and a new methodology to capture, storage and present the images of the eye in order to facilitate the equipment operation, as well as the diagnosis process and medical report, being consonant with telemedicine demands. Aiming to achieve these goals we will continue the research initiated by our team, which has succeeded in creating a proof of concept for the product we are proposing. We have worked on the basis of an optical architecture to allow the illumination and imaging of the retina without the use of polarizers to control unwanted optical artifacts, illuminating homogeneously the eye fundus, with low optical power, ensuring comfort and safety to patients, and a satisfactory resolution for medical imaging. Our proof of concept has generated a series of future demands and challenges, which we intend to address through the research and development of new computational algorithms to treat and automate some processes involved in medical examinations, by using Computer Vision techniques combined with Embedded Electronics. This will be used to perform the segmentation of the retinal arteries and veins and to recognize the level of homogeneity of the illumination, allowing the construction of retinal panoramic images simultaneously with the operation of the equipment. We also try to go further in the research and development of the optical system, optimizing their performance with the aid of new software and computational resources, as well as the construction of bench tests for optical calibration. In addition, we will integrate our algorithms programming them on mobile devices (iOS and Android platforms), building applications able to control the parameters of the cameras, with graphic UI, generation of medical report, sharing the examinations and the possibility of exporting them to a remote database. The hand-held portability we are proposing will facilitate the use of this technology away from large urban centers, and this solution also appears as the most suitable option when it is necessary to capture retina images from infants, bedridden patients or mentally handicapped. In this condition, the current table-top fundus cameras are used in a poor way. Ophthalmologists with limited investment capabilities or with little physical space in their clinics may have easier access to this kind of equipment, increasing the number of professionals working in the area. This technology tends to greatly increase the habit of these professionals to record images from the retina of their patients, generating benefits for all people involved in visual health. (AU)